Development of a 0/1D electro-thermal battery model for electric vehicles

Battery electric vehicles (BEVs) provide practical solutions to mitigate greenhouse gas emissions and decrease reliance on fossil fuels. These vehicles produce no tailpipe emissions and operate with high efficiency, particularly when using renewable energy sources for battery charging. Among the drawbacks of electric vehicles, one of the most significant is the storage capability of batteries and the consequent issue of vehicle range. In this regard, lithium-ion (Li-Ion) batteries nowadays have the highest specific energy and specific power, ensuring high average charging currents with excellent cycle endurance. However, the temperature sensitivity of these batteries is a significant challenge, underscoring the need for efficient battery thermal management systems. In this context, it is extremely important to focus on battery design, control optimization, and the development of reliable and fast models to simulate battery performance under dynamic operating conditions. In such a framework, the work carried out within the thermal and fluid dynamics department of FPT Industrial S.p.A. had the purpose of developing and validating, using GT-Suite software, a 0/1D electro-thermal model of a battery pack for light-duty vehicles. Beginning with the 3D CAD representation of the battery, the thermal model was developed using cell-to-cell discretization, followed by the creation of an equivalent electrical model. Calibration of this electrical model was conducted at the cell level, for a wide range of temperatures and C-rates. The calibration was then extended to the module sub-unit and subsequently to the entire battery pack. The electro-thermal model was validated with experimental data, first at the module sub-unit level for various C-rates, and then for the entire battery pack. Since it is required to have a reliable and fast model for control and design optimization purposes, it is decided to reduce the cell-to-cell discretization model (slower than real-time) to a less detailed module-to-module discretization one (about 6 times faster than real-time). The results obtained from the detailed electro-thermal battery pack model were compared with those from the less detailed one that employed module-to-module discretization. In the final stages, both electro-thermal models were integrated into a predictive vehicle model and subjected to testing using the Worldwide Harmonized Light Vehicles Test Procedure (WLTP). A comparative analysis of the results from the two models was performed, shedding light on the impact of the chosen level of detail in the battery model on the performance outcomes within vehicle system simulations.